Vox NT50H bias issue

[Mars Amp Repair]

Hello all,
I've been working on one of these HT50 amps that has an issue with one tube red-plating. I see where the bias is fed from the bias adjustment controls, thru the master volume control Vr1a & b. The bias is not affected by the position of the master vol pot as the signal ground side of the pot goes to ground thru a 22uf cap to ground, so the DC from the bias is isolated (BTW, I replaced that cap C16 because it was easy to do. It wasn't bad).
On this particular amp, the mstr vol control does affect the one EL34 by dragging the bias down to abt -12v as it is rotated cw. BUT, I've lifted the coupling cap to the PI, C23, and the voltage on that side of the mstr vol remains at -12Vdc. The other half of the mstr vol behaves normally maintaining the nominal -30V bias voltage throughout the entire rotation.
I don't find any inordinate resistance measurement differences between the two signal paths from the PI to the output grids.
NOTE: I haven't pulled the board to see if anything weird is going on there. If you've worked on these beasts, you'll know why!
That might undoubtedly be the next step.
The customer indicated that he had done some research, and the consensus is that there is an issue with the mst vol pot. I could believe that, if I could find a definitely reason to do it based on resistance readings or SOME indication it's bad. Guess I'm just being stubborn...
Anyone had this issue?
Thanx, Glen (JUST PULL THE DAMN BOARD!)

[Helmholtz]

Polarity of C16 correct (i.e. positive to ground)?

[The Dude]

I'm sure you probably have, but didn't say so. Have you swapped tube positions to see if the problem follows the tube or not? It could just be tube leakage.

[Jazz P Bass]

Check that the master control reaches ground when turned full off.

[Mars Amp Repair]

Thanx guys,
to The Dudes comment: I forgot to mention that I am testing the unit with the power tubes removed to prevent the one that is red-plating from being damaged. Removing the PI changes nothing.
To Helholtz: The cap is in correctly.
To Jazz P: The pot is only signal connected to ground to shunt the signal and not the bias that flows thru it. The pot does resistance check properly.
So, Since I disconnected everything from the mstr vol pot and still had -12Vdc on the fully CCW of the pot, I removed the mounting nut & thereby isolated the pot from the chassis.
Apparently, the pot has internal leakage to its case that is causing the bias issue! This sounds a lot like the issue Fishman had with their pots having shards of metal inside doing the same thing. Their solution is to blow compressed air in the pots to clear whatever the metal is in them. I did that first to no avail. I should have continued to pursue that idea!
...and guess what? I STILL HAVE TO REMOVE THE DANGED BOARD!!
Thanx all for the suggestions.
I'll be removing the dual ganged pot & trying to see what it actually is in there that is causing the issue. I'll follow up!
Glen

[Helmholtz]

Apparently, the pot has internal leakage to its case that is causing the bias issue!

What does your Ohmmeter tell you?

[Mars Amp Repair]

What does your Ohmmeter tell you?

Hello Helmholtz,
The readings don't make any sense to the case of the pot. I'm thinking there may be something going on on the top side of the board, as I'm getting readings of about 20K to the case of the pot from any of the leads of ether gang to the case.
Well, I fired up my compressor to get some more umph out of it than the canned air provides, & just blasted the pot thru the opening in the rear of the pot and the little square opening in the side of it, and voila, the resistance to the case is gone.
This IS just like the Fishman pots in their SA220 that end up with metal shavings or something inside them that short the pot to the case. Fishman's solution is to blow compressed air in the pots to remove the debris. If that doesn't work, they then suggest replacing the pots.
I'll mess around with it a bit more to see if any resistance returns and probably not replace the potif it seems stable.
As always, thanx for everyone's input.
Glen